Turkish-Swedish Bilingual Third Language English High-Front Vowel Category Formation

Abstract: This study explores the possibilities of phonetic category formation in early bilingual Turkish-Swedish learners of English. Specifically, the high-front unrounded vowels across the three languages Turkish, Swedish and English are investigated. The bilinguals are compared to L2 learners of English with Turkish and Swedish as their first language, respectively, to aim to see if the English vowel categories /i/ and /ɪ/ would be harder to establish, since the bilinguals already have three similar categories across two languages. It is hypothesized that if the bilinguals have managed to keep the Turkish and Swedish categories separate, it will have made it more difficult to establish new categories for English, since having a larger phonological inventory is thought to increase the likelihood of equivalence classification in subsequent learning. The results reveal that all three groups of speakers produced English /ɪ/ similarly to one another, but the L1 Swedish speakers made the most consistent distinction of English /ɪ/ compared to the other vowels. Furthermore, the bilinguals produced the Swedish long allophone [iː] markedly differently than the monolingual Swedish speakers. The bilinguals’ categories for Turkish /i/, English /i/ and Swedish [ɪ] had merged except in one speaker, who produced Turkish /i/ slightly further back than the Swedish short allophone. This speaker had not established a new category for English /ɪ/, but since several of the bilinguals who had merged categories had not established a new category for English /ɪ/ either, the hypothesis is not directly supported. In other words, since L2 Swedish [ɪ] had merged with L1 Turkish /i/ for almost all bilinguals, this study does not provide evidence that bilinguals’ increased number of phonetic categories across two languages makes it easier or more difficult to establish new categories for an L3.